The HIV/AIDS epidemic is one of the biggest social, economic, and scientific challenges of our time. More than 35 million people are currently infected with HIV, but only about half of these individuals are on highly active antiretroviral therapy (HAART). While the field has come a long way and saved many lives through HAART, we must now find a functional cure for infected individuals and develop an effective HIV vaccine to protect those at risk. The main barrier to complete elimination of HIV is a small proportion of latently infected cells that harbor a transcriptionally silent but stably integrated HIV genome, thus allowing HIV to evade the immune system and HAART. Long-lived resting memory CD4 T cells residing in both lymphoid and non-lymphoid peripheral tissues have been identified as the main reservoir for these latent cells; however, the mechanism of how latency is formed in vivo remains unknown. The most widely accepted hypothesis posits that memory CD4 T cells become infected during the transition from activated effector to resting memory cell, the latter being much less permissive for HIV gene expression due to profound differences in the transcriptional environment. These findings have directed the field to focus on discovering methods to specifically target and reactivate this latently infected pool in the hope of purging th virus and infected cells from the immune system. In this proposal, I challenge this model and propose that although resting memory T cells may represent a fraction of this latent reservoir, anergic T cells make up the majority. This is a significant distinction, as memory T cells have a low threshold for re-activation, while anergic T cells are non-responsive and require very strong signals to reactivate. We will test the hypothesis that latency is an accidental consequence of improper innate immune activation by HIV, occurring in HIV-specific T cells that attempt to activate under sub-optimal conditions but instead become anergic and thereby serve as an ideal host for latent HIV. We will use a newly developed viral latency reporter construct and a new genetically modified and highly improved humanized mouse model, in which several growth factors have been replaced by their human counterparts, to develop a model of in vivo latency. We will utilize human induced pluripotent stem cells (hiPSCs) generated from HIV elite controllers and non-controllers to differentiate CD34 hematopoietic precursor cells in vitro. These resulting cells will be engrafted into our new humanized mouse model that expresses the appropriate human HLA. These humanized mice are expected to recapitulate the immune systems of HIV elite controllers and non-controllers, thus allowing us to dissect the early events after mucosal HIV transmission and latency formation. We will also investigate the distribution and functionality of latent cells in vivo and identify the pathways that are involved in maintainin their particular niche. Finally, we will use TALEN technology in iPSCs to swap out protective HLAs to identify other pathways that contribute to formation of latency and elite control of HIV.